There are so many choices in AI models right now that it can make your head spin. In today's Complete Conversion System wrap-up with my group of very smart and curious students, I explored some of my workflows and why I use different models for different tasks.

But first, let’s talk about the new release from Open AI - ChatGPT 4.5, which should have become available to paying subscribers today.

Level up your communication

Last year, in many of my keynotes, I talked about Communications Quotient (CQ) - yes, it’s a term I made up to sit alongside all the other “Q’s” - IQ, EQ and AQ.

In an attempt at a little humour, this was one of my favourite slides (and who doesn’t love a Who Wants to be a Millionaire meme)?

The answer, of course, is all of the answers are correct, with Answer (A) being the simplest - (and likely least personal) - and then increasing in degree of difficulty from there on with the communication becoming more tailored to the person and the circumstances.

Answer B is the most interesting one; and is actually is the best way I can describe ChatGPT 4.5 to you.

Basically, it’s ChatGPT 4.0 but with a higher EQ.

Unlike previous models, which sometimes felt robotic, blocky, and lacking in personality, GPT 4.5 brings genuine warmth and personality to writing.

It understands emotional nuances - something other AIs wouldn't think to add.

Here is a bit of a silly example I would use to demonstrate the roboticness of the usual answer from ChatGPT compared to something with a high EQ.

I put this prompt into ChatGPT 4.5 this afternoon, and here is what I got (no need to ask for the high EQ!)

As you can see, there is a lot of EQ in this response (maybe even a little bit too much!)

ChatGPT 4.5 compared to all the other models

For me, I choose a different AI model that works with the task - whether it be writing, researching, brainstorming, or analysing data.

• For writing emails and client communications, listing descriptions, ChatGPT 4.5 or Claude 3.7 Sonnet would be my picks. To me (and this is entirely subjective), they seem to have the best writing style - and now emotional intelligence

• For research on market trends, neighborhood stats, and property histories, use ChatGPT with Deep Research or Perplexity - they can search the web for current data.

• For number crunching and financial analysis, try a reasoning model like DeepSeek R1 or Claude 3.5 Opus in ‘thinking’ mode (Anthropic’s term for reasoning).

• For coding, try Claude 3.7 Sonnet (I made a whole lot of changes to the look and feel of eliteagent.com on the weekend using Screenshots and Claude…)

Bottom line: Start by identifying what you're trying to do before opening any AI tool, then pick the one designed for that task. Using the right tool will cut your work time in half and improve your results.

Best AI Models For Common Real Estate Tasks

Again, this is subjective, but based on what I’ve found so far in my testing, if I were a real estate agent, here’s the model I would use for each purpose (assuming no budget constraints!)

Writing & Communication

For creating property descriptions, client emails, and marketing materials:

• ChatGPT 4.5: Best for warm, human-sounding writing with emotional appeal

• Claude 3.7 Sonnet: Excellent for sophisticated, polished writing styles

• What to choose: Use GPT 4.5 when you want content that connects emotionally with clients; use Claude when you need more professional, detailed descriptions

Research & Information

For market analysis, neighborhood research, and staying current:

• ChatGPT with Deep Research: Now available to all ChatGPT Plus subscribers, this powerful tool searches across the internet for current information

• Perplexity: Offers similar research capabilities with a free tier

• Grok: Good for for very recent information market and economic updates as it’s linked into X (Twitter) Feeds

• What to choose: Use Deep Research for thorough research projects, Perplexity for quick lookups, and Grok when you need something up-to-the-minute.

Brainstorming & Creativity

For marketing ideas, business strategies, and problem-solving:

• ChatGPT 4.5: Stands out for creative brainstorming and idea generation

• ChatGPT o1: For creating and validating prompts for use with the Deep Research models

• Claude 3.7: Good alternative with different creative approaches

• What to choose: Use one of the ChatGPTs when you need fresh marketing angles or business strategies; Claude when you want to refine existing ideas

Practical Tips For Real Estate Agents

1. Create task-specific bookmarks: Set up browser bookmarks for each AI tool labeled by task (e.g., "Property Descriptions - GPT 4.5", "Market Research - Deep Research")

2. Test different models for the same task: Try having both ChatGPT and Claude write the same listing description and see which better matches your style (everything is a test until it isn’t)

3. Consider the costs:

   • ChatGPT Pro: $200/month

   • ChatGPT Plus: $20/month for GPT 4.5 and 10 Deep Research searches

   • Claude: Free basic access, paid plans for more features (around $20 per month)

   • Perplexity: Free tier available, Pro plan at $20/month, gets you access to all the models

   • Grok: Free, but to level up with “Super Grok” it’s $30/month

   • Remember that saving an hour per day easily justifies these costs

4. Build a prompt library: Create and save effective prompts for each tool for repeatable tasks like (I use Notion and Blaze Text for these purposes!)

   • Property description templates

   • Client follow-up email frameworks

   • Competitive market analysis requests

5. Combine their strengths: Use Perplexity to research a neighborhood, then feed that information to GPT 4.5 to create an emotionally engaging listing description. Or feed it into Google Notebook LM to create a podcast.

   The small learning curve of using multiple AI tools pays off quickly in time saved and better quality and more interesting output.

Getting Started With Multiple AI Models

If you've only been using one AI tool, here's how to start taking advantage of their different strengths:

1. Start with the big two: ChatGPT Plus and Claude free tier will cover most of your needs

2. Add specialised tools as needed:

   • If you do lots of market research, add Perplexity

   • If you do complex financial analysis, try DeepSeek

   • If you need to stay on top of trends and like X, consider Grok

3. Track your results: Notice which AI gives you better results for specific tasks and refine your system

4. Stay flexible: These models are improving rapidly, so what works best today might change next month

5. Share with your team: Get your office AI Champion to make sure everyone in your office knows which tool works best for which task.

All of these models are improving at lightning speed, so staying aware of their changing capabilities will give you an edge over agents who set and forget their AI workflow. As I find myself saying quite often these days, today is the worst any of these models will ever be…

🛑 Don’t forget the golden rules!

No matter which model you use:

• It’s always a first draft

• There should always be a human in the loop

• Always do a fact check - it doesn’t matter which model they can all still hallucinate

Using the right AI for each task isn't just about efficiency—it's about creating better outcomes for your clients.

As we move into the era of Agentic AI, it’s time to start thinking of your AI tools like a team of assistants, each with their specialty, rather than trying to find one that does everything perfectly.

Which is your favourite model?

Happy hunting 🚀

—Sam

Co-Founder of Elite Agent

Creator of 30 days of Practical Property Prompts

Creator of The Complete Conversion System 

Have a question about this post? Join the Community